Application of the Budyko Framework and the Complementary Relationship to Estimate Evapotranspiration.

Abstract:

Recently, Anayah and Kaluachchi (2014), proposed an improved version to the Ganger and Gray (GG) method that uses the complementary relationship and local meteorological data only to compute regional evapotranspiration (ET). This modified model showed excellent performance compared to other published studies but the method needs improvements under dry conditions. To overcome this deficiency, we propose a further refinement to the earlier work of Anayah and Kaluarachchi (2014) using both the complementary method and the Budyko framework. The difference is the addition of the Budyko framework which can capture vegetation cover better than with the complementary method only. This work used 75 AmeriFlux eddy covariance tower sites in USA. The proposed model called GG-NDVI reduced the mean root-mean-squared-error (E_RMS) by 32% and 15% for dry and wet regions, respectively, as compared the modified GG method. From the comparison with other complementary methods, the GG-NDVI method showed the lowest mean E_RMS of 14.7 mm/month for the dry regions. Most dry regions underestimated ET during the summer seasons. We also observed some decrease in E_RMS in the wet regions with increasing precipitation and with higher plant density associated with high ET. While the GG-NDVI method showed good performance in most regions, a better parameter than NDVI is needed to accommodate regions with dense vegetation cover.